European green crab carapace width is a measurement, not an automatic age certificate. A crab measuring 70 millimetres across the shell is larger than one measuring 25 millimetres, but the difference does not tell us when either animal hatched or where it originated. Growth depends on moulting, conditions and individual history. An accurate translation preserves what was actually measured before attempting to explain biological meaning.
This guide teaches readers to interpret European green crab size-frequency charts, sex observations, egg-bearing females and possible recruitment signals. Every calculation below uses explicitly hypothetical teaching data, not newly reported observations from a real site. The task is to distinguish measured facts from inferred processes, and to give readers the numbers, denominators and cautions they need to verify an explanation.
The quick method is to confirm species and units; define how carapace width was measured; distinguish captured from measured animals; retain unknown-sex and missing-width records; show the sample size for each percentage; state the boundaries of size classes; compare matched gear and seasons; and describe recent recruitment as a possible explanation when supporting evidence remains incomplete.
Measurement first, biological inference second
The carapace is the upper external shell. Measuring its width requires anatomical landmarks and instrument precision. If one laboratory measures across lateral spines while another uses a different axis, the resulting numbers may not be interchangeable even if both are recorded in millimetres. A unit conversion changes the scale of the number, not the structure it measures. Keep the physical definition attached to every exported value.
Crabs grow through moulting: a restrictive exoskeleton is replaced as the animal expands, then hardens. Intermoult periods and growth increments vary. Age cannot be calculated from width alone without extra assumptions and context-specific growth evidence. It is therefore misleading to label every small crab a specific age or every large crab a known number of years old merely by looking at one histogram.
Washington Sea Grant’s Crab Team records numbers of crabs and takes size and sex measurements on defined subsets. Its monitoring reports discuss small captures, established populations and observations of egg-bearing females. These strands of evidence are useful, but each has a different denominator and interpretation. They should not be flattened into a single unexplained claim about how quickly a population is breeding.
Suppose eighty target crabs were captured and only twenty measured because they were easiest to handle. Six among the measured animals fall below a screening width. Six divided by twenty is 30 percent of the measured subset, not necessarily 30 percent of all eighty catches. If the twenty were not selected representatively, extrapolating their size distribution to the full catch requires caution.
Denominators, sex and reproduction
Sex ratios require similar discipline. If twelve identified females, eight identified males and five unclassified crabs are captured, the female proportion among the twenty sexed animals is 12 ÷ 20 = 60 percent. Among all twenty-five captured animals, confirmed females are 12 ÷ 25 = 48 percent with five unknown. Neither fraction proves that the estuary population has precisely the same sex balance.
An egg-bearing or ovigerous female is an individual visibly carrying eggs. That observation supports a specific statement about her reproductive state. It does not by itself prove successful larval hatching, local settlement, juvenile survival or self-sustaining recruitment. A larva can be transported by water movement, and a positive trap does not reconstruct every stage of its life history.
Size classes should have unambiguous edges. A histogram that uses 20 mm to under 30 mm and 30 mm to under 40 mm places an exact 30.0-mm observation in the second class. If different analysts place boundary animals on opposite sides, the chart may show a false trend. State the convention and calculate with the original recorded precision before rounding displayed values.
Missing measurements are not zero-width crabs. An animal with damaged anatomy may have an unrecorded width; a small specimen may remain unclassified by sex. Keep ‘not measured’ and ‘unknown’ separate from biological categories. This protects an analyst from producing misleadingly exact percentages by silently dropping awkward rows.
Size-frequency observations are most persuasive when accompanied by dates, stations, gear and sampling scope. A juvenile-heavy series from one method and adult-heavy series from another could reflect sampling selectivity, not ecological difference. Good translation distinguishes the structure of an observed sample from claims about all animals in an unsampled estuary.
Seventeen worked cases: calculation, meaning and independent transfer
Case 1: An exact boundary width
The measured widths are 29.9, 30.0 and 30.1 millimetres. A graph uses intervals from 20 to under 30 and 30 to under 40. One analyst places 30.0 into the first class merely because the legend says 20–30, while another uses the second class.
Under a lower-inclusive, upper-exclusive convention, the first bin has one animal and the second has two. This is a boundary decision, not an ecological discovery. Document the class convention in the methods or figure note.
A useful correction says exactly which bin contains a boundary observation. A chart cannot be considered reproducible if different people can assign the same measurement to two bars.
Transfer exercise: place a 40.0-mm crab into 40 to under 50 rather than 30 to under 40. Count it once, preserving the original decimal precision.
Case 2: A striking maximum hides the sample
Five hypothetical widths are 25, 27, 29, 31 and 82 mm. A publicity caption reports that green crabs at the site are 82 mm wide, obscuring the four much smaller individuals in the same little sample.
The sum is 194 mm, giving mean 38.8 mm. The sorted median is 29 mm, and the range is 25–82 mm. Each statistic is valid but tells a different part of the story.
Write the sample size and distribution, not merely the largest observation. A reported maximum is not the typical animal and is not a size threshold for all crabs at the location.
Transfer exercise: the values 20, 20, 20 and 80 mm have midpoint 50 mm but mean 35 mm. Do not mistake midrange for arithmetic mean.
Case 3: A measured subset selected for convenience
A crew captures forty green crabs in several traps but measures only ten that are easy to reach. Three measured animals fall below a juvenile-screening width. A chart announces that 30 percent of the bay’s crabs are young.
The observed statistic is three divided by ten, or 30 percent of the measured subsample. The remaining thirty captured crabs have unmeasured widths. No random-sampling or representativeness evidence was provided.
Label the chart ‘widths of ten selected captures’ and state the rule used to select them. If selection favours larger or calmer animals, a population percentage would be especially unsafe.
Transfer exercise: if the programme later measures all forty and twelve are small, 12 ÷ 40 = 30 percent of captures; that still is not a complete census of the estuary.
Case 4: Rounding before measurement analysis
Widths of 34, 34 and 34 mm are entered as 30, 30 and 30 after a worker rounds to the nearest ten millimetres. The displayed mean becomes 30 mm even though the original three readings all show 34 mm.
The true sample mean from the raw values is 34 mm. Rounding measurements before computing a mean created a four-millimetre processing shift, not a biological effect.
Keep raw instrument resolution, apply quality checks, compute statistics and then round the published summary appropriately. Retrospective rounding can also move individuals across histogram boundaries.
Transfer exercise: convert 7.5 centimetres into 75 millimetres first. Do not silently compare that value with a 7.5-millimetre observation.
Case 5: The unknown-sex ledger
Survey records show twenty-five green crabs, twelve classified as female, eight as male and five not sexed. An editor reports the female proportion as 48 percent without acknowledging which denominator was used.
Twelve of twenty classified animals are female, or 60 percent of those sexed. Twelve of all twenty-five captures are known females, or 48 percent of all captures; the remaining five have no known sex assignment.
Both numbers can be useful if properly labelled. Neither number establishes the population sex ratio because capture behaviour and classification can bias the observed sample.
Transfer exercise: if two formerly unknown animals are confirmed male, the classified total rises to twenty-two, giving 12 ÷ 22 ≈ 54.5 percent female among classified crabs.
Case 6: An adult-biased capture method
A programme mostly uses one gear type that tends to catch active larger crabs and uses a second method at a few sheltered juvenile habitats. The combined histogram shows overwhelmingly large crabs.
The distribution is weighted by where and how the sample was collected, not merely by the true number of crabs of each size. Ninety adult-biased captures and ten juvenile-focused captures should not be pooled without retaining their methods.
Compare within gear categories or apply an explicitly supported model. Statements about the whole community must account for the sampling design and unsampled life stages.
Transfer exercise: a single small individual found in an adult-biased trap proves that small individual was present, not that the trap has high sensitivity to all juveniles.
Case 7: A small-size pulse in autumn
At one station no crab under 25 mm was captured in an early survey. Fourteen such crabs appear in a late-season sampling event. A story claims all fourteen were born locally in the same recent breeding episode.
An autumn concentration of small captures may be compatible with recent recruitment to the habitat. Yet larvae may have arrived with currents and individuals may differ in growth rates, activity and availability to traps.
Explain the place, date, threshold, counts and changed effort. A biologically plausible interpretation should be labelled as such rather than becoming an invented history of fourteen individual animals.
Transfer exercise: if fourteen small crabs are found at one of ten stations, do not write that all ten stations experienced the same recruitment pulse.
Case 8: A visible brood on one female
An egg-bearing green crab female appears among forty confirmed captures. The photograph documents a real observation, while a proposed press release calls it proof that an extensive breeding population is already established.
The observed fraction is one egg-bearing capture out of forty target captures, or 2.5 percent of that catch. It is not a 2.5 percent population ovigerous rate without relevant female denominators and capture selectivity.
Local reproduction, hatching and successful recruitment are different biological stages. Describe the egg-bearing record accurately, then identify the additional evidence needed for claims about population persistence.
Transfer exercise: three ovigerous females among thirty female captures is 10 percent of captured females, not 10 percent of all green crabs in the bay.
Case 9: Comparing different seasons
Thirty widths are recorded during a spring visit in Year One and fifty during an autumn visit in Year Two. The second histogram contains more small individuals. The accompanying headline attributes the entire difference to greater reproduction.
The comparison changes both time of year and sample size. Seasonal growth, availability, gear conditions and station choice may influence the result. A credible annual trend needs matched windows and documented protocol stability.
Report the higher observed small-width share in the measured autumn sample, then explain that a year effect cannot be isolated from these two unmatched observations.
Transfer exercise: twenty of fifty is 40 percent and ten of thirty is about 33.3 percent, a difference of 6.7 percentage points, not a demonstrated 6.7 percent population increase.
Case 10: First documented capture versus arrival date
A large individual is caught in a bay after several years without any formal survey. A local article states that the crab must have arrived years ago solely because it was large.
Body size may be consistent with survival over time, but a precise arrival date cannot be reconstructed from one shell width. Animals move, larvae disperse, and past nondetections depend on monitoring effort.
Use ‘first documented capture under the current programme’ for the observation. If specialists estimate an age range using regional growth data, cite their assumptions separately.
Transfer exercise: a new record in 2026 is not necessarily the species’ first physical presence in 2026. Detection chronology and arrival chronology are different.
Case 11: A midpoint falsely named a mean
A chart only gives minimum width 20 mm and maximum width 85 mm. A writer calculates 52.5 mm and calls it the average size. There are no individual measurements available to support that claim.
Fifty-two point five is the midrange: (20 + 85) ÷ 2. The mean requires all individual widths or a reported aggregate, while the median requires knowledge of the distribution’s middle position.
Preserve the original range and refrain from constructing a mean. A statistics glossary should not allow one word ‘average’ to hide two different calculations.
Transfer exercise: with widths 20, 20, 20 and 80 mm, mean is 35 and midrange is 50. Check which measure the source actually reports.
Case 12: Two apparent modes in a histogram
A width plot has visible peaks near 25 and 65 mm. The legend instantly labels them ‘one-year-old’ and ‘three-year-old’ cohorts, although no age measurements or validated growth mixture analysis were performed.
Two peaks are observations about a binned distribution. They may reflect multiple size groups, gear, sex, habitat mixture or sampling conditions. The age-class assignment is a biological model, not a direct measurement.
Use ‘two apparent size modes’ and give sample size, bin width and gear context. Test whether the pattern remains when bins shift moderately before proposing strong cohort stories.
Transfer exercise: if doubling bin width merges the two peaks, the visual explanation is fragile. Compare actual widths rather than relying on a bar-chart impression.
Case 13: Station averages versus pooled ratios
Station A contains ninety classified females and ten males; Station B contains one female and nine males. One writer averages the station female percentages equally to say the whole combined sample is 50 percent female.
A has 90 percent and B has 10 percent female; their equal-station average is 50 percent. However, pooled captures contain 91 females in 110 classified animals, so the pooled female fraction is about 82.7 percent.
Both results can be mathematically described, but the choice of weighting changes the estimand. State whether the summary gives an equal vote to each station or to each captured individual.
Transfer exercise: do not infer that either pooled or unweighted sample automatically represents the biological sex ratio across unsampled locations.
Case 14: Width, length and centimetres in one column
A database import combines carapace width in millimetres, carapace length in millimetres and a second lab’s centimetre measurements under a generic ‘size’ field. A 7.8 entry can therefore represent radically different dimensions.
Unit normalization converts 7.8 cm to 78 mm. It does not transform a measured length into a measured width. A morphometric conversion between different axes needs a validated species- and context-relevant relationship.
Restore separate fields for anatomical measurement, numerical value, unit, instrument and protocol. A clean-looking single column is harmful when it conceals incompatible physical quantities.
Transfer exercise: 83 mm equals 8.3 cm, not 83 cm. State the unit in both the table header and public prose.
Case 15: A changing observer practice
An experienced surveyor sometimes classified uncertain small animals as male, while a later team records them as unknown sex. The percentage male drops immediately after staff turnover, inviting a dramatic demographic headline.
The change may be procedural rather than ecological. Preserve observer names, field-guide versions, photography protocols and reasons for uncertain assignment. Review archived evidence if appropriate without silently reclassifying historical records.
Report the shift in classification practice alongside the observed proportion. Quality-control transparency can prevent a false story about a sudden change in population reproduction.
Transfer exercise: a rise in unknown sex from zero to 20 percent after retraining is not evidence that 20 percent of crabs acquired a new biological sex.
Case 16: Size subsample versus total removals
A removal crew captures one hundred confirmed green crabs but has time to measure only twenty. The annual summary mistakenly reports twenty as the total captured because the size-analysis file was used as the master catch ledger.
One hundred is the capture count; twenty is the size subsample. The two quantities are related but distinct. If twelve of twenty measured are female, this describes the measured subset subject to sex coverage, not sixty of the hundred.
Retain a raw capture table and a linked individual-measurements table with stable specimen IDs. A join by record ID avoids replacing one denominator with another.
Transfer exercise: if a chart says n = 20 widths, do not read it as total trap capture unless that is independently confirmed.
Case 17: An ecological conclusion from repeated classes
A cove yields small crabs in one year, larger crabs in the next and an occasional egg-bearing female. These multiple observations are more informative than a single capture, but their combination still cannot identify the exact birthplace of larvae.
Consistent repeated observations can support concern about sustained local presence and reproduction. Migration, growth and larval transport remain possible. Compare matched monitoring records and independent settlement evidence before declaring local self-replenishment.
A good synthesis identifies each observation, its timing, method and uncertainty. Words like ‘consistent with’ make the evidence stronger by describing exactly how far it reaches.
Transfer exercise: one season without small captures could reflect missed juveniles, altered timing or low recruitment; it does not establish that all reproduction stopped.
Further field cases: interpret the measurements, not the assumptions
Further case 1: Molt remains are not measured live crabs
A team finds three identifiable shell molts on the shoreline and captures five live crabs in traps. The published table places all eight into the live-crab size histogram because the molts can also be measured.
Molts are evidence of discarded exoskeletons and should be recorded separately from live captures. They may have moved since shedding; one animal can leave more than one molt over time, so an exoskeleton is not an unambiguous independent living individual.
Measure molts using their documented method if scientifically useful, but keep separate labels and sample sizes. A molt width is not a live specimen’s current exact size after the molt.
Transfer: five live crabs and three molts are five live captures plus three molt observations, not eight independent living crabs.
Further case 2: A larger mean can come from site weighting
One marsh yields ten measured small crabs in spring. Another yields ninety larger crabs in summer. A pooled mean rises when the second marsh is sampled more intensively, even if individual site distributions remain unchanged.
The mean of pooled observations weights the second station nine times more heavily. Distinguish a change within comparable stations from a change in which stations supplied observations. Site composition is a cause of apparent overall shifts.
A public explanation should show station-specific sizes and sample sizes rather than only the combined mean. The choice of estimator is part of the biological question.
Transfer: equal weighting of two station averages is not identical to a weighted mean of every measured crab when n differs.
Further case 3: Uncertainty of a single reading
A calliper measures a shell width at 42.3 mm, but repeated placement differs by approximately 0.5 mm. One chart reports a value to three decimal places and makes a sharp cohort boundary at 42.300 mm.
The extra decimal places are not supported by the repeatability of measurement. Record the instrument reading according to the field protocol and describe plausible measurement uncertainty when it affects class assignment.
If the observed width sits close to a bin boundary, repeat the measure or flag ambiguous classification instead of assigning false precision. Biological conclusions should not depend on a tenth of a millimetre unsupported by technique.
Transfer: a reading of 39.8 mm with possible variation of about 0.5 mm can cross a 40-mm class edge; note this when interpreting a narrow bin.
Further case 4: Manual versus photograph-based size
A survey records calliper measurements for one year and estimates widths from photographs in another. Photos are attractive for remote review, yet scaling, camera angle and specimen posture can introduce measurement differences.
Assess whether each image contains a reliable scale, whether perspective is controlled, and whether the photo-based method has been validated against direct widths. Without that information, distributions may not be directly interchangeable.
Name the measurement method in each year and avoid portraying a technical switch as a biological trend. A cross-calibration dataset can help estimate the method difference.
Transfer: a twenty-percent apparent change in mean width after switching instruments should trigger a measurement audit before a demographic conclusion.
Further case 5: Mature is not synonymous with large
A field guide calls a certain width class adult-sized. An editor rewrites this as ‘sexually mature’ for all animals above the class threshold, without a reproductive maturity assessment.
Size and maturation may be associated, but a threshold designed for convenient reporting is not automatically a verified reproductive criterion in every habitat. Maturity depends on biology and sometimes individual condition.
Use ‘larger width class’ or the exact reporting category unless a tested maturity criterion is part of the source. Do not turn a screening bin into an individual medical-style diagnosis.
Transfer: a large male’s width does not demonstrate that it successfully reproduced in the sampled season.
Further case 6: A juvenile label changes across datasets
One project defines a small-crab class below 30 mm; another uses below 35 mm for a management screen. Both tables use the word juvenile, and a translation merges the percentages as if their thresholds matched.
These are nonidentical operational categories. Recalculate from original widths under a common declared class if the data permit. If raw measurements are unavailable, preserve each source’s threshold instead.
The correct text reports ‘under 30 mm’ or ‘under 35 mm’ explicitly rather than relying on an apparently universal label. Ecology is not improved by concealing classification rules.
Transfer: a 32-mm crab belongs to the first project’s larger group but the second project’s small group. Neither assignment alters its actual measured width.
Further case 7: A sex ratio in one trap does not fill the shoreline
At one active tidal channel, ten males and two females enter baited traps. A newspaper headline announces that the whole shoreline contains five males for every female.
The catch ratio at the station is 10:2, or 5:1, but it is conditional on gear, time and local movement. Spatially separated females or differential response to bait may alter the proportion.
Translate it as the captured ratio for that method and station. To infer coast-wide sex ratios, the study needs representative spatial sampling and species-specific detection assumptions.
Transfer: a second nearby trap finding two males and ten females demonstrates that local samples can differ; pooling should retain effort and geography.
Further case 8: One unusual sampling day
A crew measures sixty crabs on a day after unusual weather. Only fifteen are measured at the same station during several typical visits. The storm-day values dominate a pooled monthly chart.
Measurement count weighting may reflect event opportunity rather than usual distribution. Weather can change accessibility and behaviour. A sensible report shows the sampling dates and possibly day-specific results.
Do not discard observations because they are inconvenient; record conditions and examine how results change under transparently defined sensitivity analyses. An unusual field day can be biologically interesting without representing the whole month.
Transfer: two dates with fifty and five measured crabs should not automatically be treated as equally informative or as identically sampled days.
Further case 9: An egg-bearing denominator error
Six egg-bearing females are recorded among thirty classified females in a sampling season. A caption reports six divided by one hundred target crabs, because one hundred were captured overall.
Both rates can be calculated, but they answer different questions: 6 ÷ 30 = 20 percent of captured classified females were egg-bearing, while 6 ÷ 100 = 6 percent of all target captures were egg-bearing females.
Write the female-specific percentage only with its female denominator. Neither statistic represents the underlying population frequency without accounting for selective capture and season.
Transfer: do not call 20 percent an egg hatching success rate; the numerator documents egg-bearing females, not successfully hatched broods.
Further case 10: Width uncertainty across recording languages
An international collaboration records ‘CW 45’ in one worksheet, ‘shell size 4.5’ in another and ‘width of back 45’ in translated notes. The terms look related but may not share units or anatomical definitions.
Create a glossary linking each field to a measurement protocol. Confirm whether 4.5 is centimetres, and whether shell size refers to the same width axis. Translation should preserve source ambiguity until verification, not invent certainty.
A strong bilingual data dictionary gives the original label, standard English term, unit, axis and method reference. This prevents apparently tiny animals appearing when centimetres are misread as millimetres.
Transfer: the correct conversion is 4.5 cm = 45 mm, but only after confirming that 4.5 truly represented centimetres.
Further case 11: Age class inferred from a growth curve
A local model proposes an age range for crabs of a given width, but a press summary reports an exact birthday based on the centre of that range. The model’s uncertainty disappears.
A model-derived range is an estimate conditional on fitted growth, region and time. Individual growth can vary. Even an excellent model does not transform a size measurement into a calendar birth record.
State the estimate as conditional and retain the range and method. If the growth evidence comes from another region, disclose the transfer limitation rather than concealing it.
Transfer: ‘consistent with an older size class’ is not equivalent to ‘hatched in 2023’ without a justified calibration.
Further case 12: A confirmed species is essential
A mixed catch contains green crabs and native crabs. Two specimens have ambiguous identification and are dropped into the target species size chart because their shells seem similar.
Species classification is a prerequisite for species-specific size analysis. Uncertain identifications need expert checking, photographs or other procedures appropriate to the programme. Arithmetic cannot repair wrongly attributed individuals.
Keep a separate unidentified category until review. Do not claim a population’s size structure based on measurements of a mixture of different species.
Transfer: removing two disputed specimens from a ten-animal green-crab size sample changes n to eight; update every fraction and chart label.
Further case 13: A new station changes the network
An estuary network adds a nursery habitat with many small trapped crabs. The regional mean width falls, and the annual report calls it a coast-wide shift toward younger animals.
The changed site mix can explain the result without a comparable shift at earlier stations. Compute a matched-site series first and describe the new site as expanded spatial coverage.
Large sample sizes do not remove sampling-frame bias. A network-level summary needs an account of sites present in each year, not just a pooled average.
Transfer: a station added in Year Two cannot supply a paired Year One value unless a historical dataset genuinely exists.
Further case 14: Sex assignment and body condition
A group of heavily fouled or damaged specimens has incomplete abdominal features, making identification uncertain. A clerk assigns them all to the visually dominant sex and deletes the unknown category.
That procedure manufactures confidence and can bias a sex-ratio trend. Note specimen condition and identification confidence. A review decision should be documented with the evidence and method used.
Public communication may state that some animals could not be sexed reliably. Unknown is a valid data status, not a failure to fill in a spreadsheet.
Transfer: if damaged individuals cluster in one size class, the known-sex sample may be systematically unrepresentative of that class.
Further case 15: A disappearance of small captured animals
A habitat yields many small captures one year but no individuals under the same width threshold the next. The monitoring record is important, but the report immediately declares that recruitment has ended.
The absence of small captures could reflect low arrival, juvenile survival, gear and timing differences or incomplete detection. Distinguish ‘not captured’ from ‘not present’, especially if the programme has altered sampling effort.
A careful follow-up asks whether the small-animal sampling method, season, habitat access and measurement thresholds stayed constant. Then it proposes targeted monitoring to test competing explanations.
Transfer: zero small captures among forty adults does not have the same sensitivity as zero among four mixed-size captures.
Further case 16: A defensible multi-year synthesis
Several seasons of measurements include small, intermediate and large green crabs, both sexes and occasional egg-bearing individuals. The mixed evidence is compatible with persistence and possibly recruitment in the region.
The strongest claim is determined by the design: repeated detections document ongoing occurrence; size classes describe sampled structure; egg-bearing animals document reproductive state. Local birth and survival are additional hypotheses.
Write each finding with dates and sample sizes, then explain the remaining assumptions. This produces a more credible synthesis than announcing a perfectly reconstructed local population history.
Transfer: a reader should be able to trace every noun in a conclusion—juveniles, adults, females, recruits—to a defined observation or explicitly qualified inference.
Independent checks, FAQs and decision language
Independent worked check: twenty-two widths are measured and seven are below the defined screening class. The measured fraction is 7 ÷ 22 ≈ 31.8 percent. If eighty-nine crabs were captured overall, the report must show that only twenty-two were measured and that the proportion applies to those measured. Without representative selection, it cannot be called the true estuary-wide percentage.
Independent worked check: 38 male, 31 female and 11 unknown-sex individuals sum to eighty. Among known-sex captures, the female percentage is 31 ÷ (38 + 31) = 31 ÷ 69 ≈ 44.9 percent. Checking the addition is as important as checking the division; a phrase saying ‘seventy were sexed’ would silently introduce an erroneous denominator.
Independent worked check: three stations measured ten, fifty and forty crabs. Their small-width fractions were 50, 10 and 20 percent. Equal-weight station average is (50 + 10 + 20) ÷ 3 ≈ 26.7 percent; pooled fraction is (5 + 5 + 8) ÷ 100 = 18 percent. State what is averaged and why before presenting either value.
How should a width plot be translated? Preserve its axis units, class edges, total n and sample selection. If only a graph is available, do not invent exact means or ages that cannot be recovered from the published bins.
Do small crabs prove local breeding? No. They can provide evidence consistent with recruitment into the sampled habitat, but larvae may have arrived from elsewhere and trap sensitivity may differ across stages. Breeding, settling and being captured are separate events.
Can a visible egg mass establish local population persistence? It is an observation of an egg-bearing female. Successful hatching, larval development, local settlement and multi-year persistence require other evidence. A cautious sentence can be informative without overclaiming.
How do sex ratios differ from sex ratios in traps? The latter use observed caught and classified individuals. A population sex ratio is a wider biological target and may require a representative sampling model accounting for behaviour and capture probability.
Why preserve the missing values? Unknown classifications identify where evidence stops. Deleting them changes denominators, can bias inference and prevents later reviewers from understanding why the sex or size distribution may not represent all captures.
Which statistics are safe for a small size sample? Counts, range and median are usually straightforward to compute when raw values exist, though none alone eliminates sampling uncertainty. Means and class percentages are valid descriptive measures when labelled with the sample and units.
How do you repair an overconfident English sentence? Replace ‘the population is three years old’ with the measured width and a qualified model estimate if available. Replace ‘breeding is proven’ with the observed life stages. Replace ‘all crabs’ with ‘crabs measured’ when that is the true denominator.
Scientific explanation is not a contest to sound certain. It is the practice of making every supported claim reproducible and identifying which next observation would test an uncertainty. In related guides, green crab trap CPUE and zero catches addresses effort, while lionfish reproductive indices and larval dispersal demonstrates a separate chain of evidence.
Methodological references
Washington Sea Grant, Male or Female? Crab Team field protocol
Washington Sea Grant, Ocean Shores green crab monitoring
Washington Sea Grant, coastal captures, age classes and size